Risky Business Wide World of Cyber: DeepSeek lobs an AI hand grenade
Feb 21, 2025
Join Alex Stamos, former CISO for Facebook and Yahoo, and Chris Krebs, ex-director of CISA, as they dive into the intriguing world of AI and its geopolitical implications. They discuss the controversial DeepSeek AI model and its potential impact on cybersecurity. The duo analyzes the skepticism surrounding AI innovations and examines the contrasting regulatory approaches in Europe and the U.S. Don't miss their insights on privacy rights, market dynamics, and the future of AI innovation amidst growing international tensions.
AI Snips
Chapters
Transcript
Episode notes
DeepSeek's Impressive LLM
- DeepSeek, a Chinese AI lab, published an impressive open-source LLM, raising concerns about its development cost and efficiency.
- The model's efficiency and claims of low training costs surprised many, leading to speculation about Chinese advancements in AI.
LinkedIn Post Tanks NVIDIA Stock
- A LinkedIn post misinterpreting DeepSeek's claims caused NVIDIA's stock to plummet, highlighting the market's volatility around AI news.
- This incident revealed the ease of manipulating Wall Street with AI information.
Bypassing Hardware Sanctions
- The assumption that DeepSeek only uses sanctioned hardware is likely false, given the availability of resources through third countries.
- Chinese companies can rent advanced hardware in countries like the UAE, bypassing sanctions.